Methods for high-throughput analysis of profiles of gene expression that assay thousands of genes simultaneously are powerful approaches for understanding and classifying cell and tissue phenotype. This includes analysis of normal pathways of cell maturation and their perturbation in transformation, the sensitivity and mechanism of response of normal and tumor cells to physiological and pharmacological agents, and modulation of tumor risk and progression by nutritional factors. However, the complex data generated by such approaches raise difficulties in analysis. We will describe some of the methods we have used in analyzing databases generated in a number of projects in our laboratories. These include: the role of k-ras mutations in colon cell transformation; the role of p21(WAF1/cip1) in intestinal tumor formation and response to sulindac; the development of the absorptive and goblet cell lineages; sensitivity of colonic cells to chemotherapeutic agents; mechanisms that regulate c-myc expression utilizing novel methods of transcriptional imaging; and interaction of nutritional and genetic factors in modulation of intestinal tumor formation.